Stockfish Testing Queue

Finished - 19396 tests

09-08-17 fa tuned diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 19433 W: 2481 L: 2578 D: 14374
sprt @ 60+0.6 th 1 Try LTC, since King's parameters are TC dependent: tuning some king's parameters (Low throughput)
11-08-17 ia lever_count4 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9513 W: 1671 L: 1756 D: 6086
sprt @ 10+0.1 th 1 lever count malus. Take 4
11-08-17 Gu passed diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 1543 W: 256 L: 413 D: 874
sprt @ 10+0.1 th 1 take 2
11-08-17 mc ord diff
LLR: -0.20 (-2.94,2.94) [-3.00,1.00]
Total: 966 W: 166 L: 178 D: 622
sprt @ 10+0.1 th 1 Score evasions captures as normal captures
10-08-17 Vo skipResearch3 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 58392 W: 10527 L: 10405 D: 37460
sprt @ 10+0.1 th 1 Another shot at this idea...
11-08-17 Gu materialtweak diff
LLR: -2.95 (-2.94,2.94) [0.00,4.00]
Total: 33128 W: 6015 L: 6058 D: 21055
sprt @ 10+0.1 th 1 material tweak (More significant weights of pieces)
10-08-17 ia phalanx2 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 33509 W: 5951 L: 5935 D: 21623
sprt @ 10+0.1 th 1 tweak phalanx bonus. Take 2
10-08-17 ia sort_captures6 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 12248 W: 2132 L: 2205 D: 7911
sprt @ 10+0.1 th 1 prefer captures close to back ranks
10-08-17 vd scoreSimp diff
LLR: 2.94 (-2.94,2.94) [-3.00,1.00]
Total: 34641 W: 4521 L: 4419 D: 25701
sprt @ 60+0.6 th 1 ltc, take 2
10-08-17 Gu PassedPair6 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5147 W: 860 L: 963 D: 3324
sprt @ 10+0.1 th 1 take 6
11-08-17 Gu passed diff
LLR: -2.94 (-2.94,2.94) [0.00,4.00]
Total: 1238 W: 183 L: 339 D: 716
sprt @ 10+0.1 th 1 passed tweak
10-08-17 II STCfish diff
LLR: -2.97 (-2.94,2.94) [0.00,5.00]
Total: 17623 W: 3123 L: 3174 D: 11326
sprt @ 10+0.1 th 1 I understand that this is not commitable, but I collected some "good" patches from July and August, and I want to test with low throughput do we left something to garbage collectors.
10-08-17 ia phalanx diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 25588 W: 4546 L: 4563 D: 16479
sprt @ 10+0.1 th 1 tweak phalanx bonus
10-08-17 An cmpPvNodes diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 16626 W: 2921 L: 2976 D: 10729
sprt @ 10+0.1 th 1 Use a stricter value for counter move pruning in PvNodes
10-08-17 Gu PassedPair diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 35743 W: 6488 L: 6461 D: 22794
sprt @ 10+0.1 th 1 take 4
10-08-17 ia connected diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 9277 W: 1596 L: 1682 D: 5999
sprt @ 10+0.1 th 1 avoid negative connected bonus
10-08-17 Gu PassedPair5 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 20774 W: 3663 L: 3701 D: 13410
sprt @ 10+0.1 th 1 take 5
10-08-17 pr ps_optimize diff
LLR: -2.02 (-2.94,2.94) [0.00,5.00]
Total: 15882 W: 2863 L: 2881 D: 10138
sprt @ 10+0.1 th 1 Slight improvement locally. VERY IMPORTANT: 10+0.1 games only.
10-08-17 Vo poorImprovement diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 10980 W: 1879 L: 1958 D: 7143
sprt @ 10+0.1 th 1 Take 2
10-08-17 sg lmr_no_opp_imp diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 9916 W: 1725 L: 1808 D: 6383
sprt @ 10+0.1 th 1 Account for tempo.
10-08-17 sg lmr_no_opp_imp diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 5598 W: 947 L: 1048 D: 3603
sprt @ 10+0.1 th 1 More LMR if opponent couldn't improve the position with his last move (corrected).
09-08-17 sg move_order diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 17090 W: 3129 L: 3181 D: 10780
sprt @ 10+0.1 th 1 Only knight moves. For them the measured correlation of distance and the chance of being good (value > alpha) is the best. (Park with prio -1)
09-08-17 sg move_order diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 33794 W: 6093 L: 6075 D: 21626
sprt @ 10+0.1 th 1 Exclude king moves
10-08-17 Gu PassedPair diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 33593 W: 5988 L: 5971 D: 21634
sprt @ 10+0.1 th 1 take 3
10-08-17 Vo poorImprovement diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 14256 W: 2497 L: 2562 D: 9197
sprt @ 10+0.1 th 1 stc
09-08-17 vd scoreSimp diff
LLR: 3.17 (-2.94,2.94) [-3.00,1.00]
Total: 70240 W: 12543 L: 12494 D: 45203
sprt @ 10+0.1 th 1 stc
10-08-17 An lmrTacticalByNonPawns diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 19154 W: 3421 L: 3465 D: 12268
sprt @ 10+0.1 th 1 reduce LMR for tactical moves by non pawns
09-08-17 Vo skipResearch2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 53881 W: 9647 L: 9545 D: 34689
sprt @ 10+0.1 th 1 take 2
09-08-17 Gu PassedPair diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 31730 W: 5724 L: 5714 D: 20292
sprt @ 10+0.1 th 1 take 2
09-08-17 ia sort_captures5 diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 40174 W: 7137 L: 7093 D: 25944
sprt @ 10+0.1 th 1 more materialistic capture sorting
09-08-17 sg king_distance_history diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 19844 W: 3467 L: 3509 D: 12868
sprt @ 10+0.1 th 1 Update weight 8
09-08-17 vd scoreSimp diff
LLR: 2.96 (-2.94,2.94) [-3.00,1.00]
Total: 27377 W: 4957 L: 4847 D: 17573
sprt @ 10+0.1 th 1 stc, take 2
09-08-17 ia pawn_file diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 29090 W: 5178 L: 5180 D: 18732
sprt @ 10+0.1 th 1 per file pawn evaluation arrays
09-08-17 sg king_distance_history diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 18407 W: 3250 L: 3297 D: 11860
sprt @ 10+0.1 th 1 Update weight 16
09-08-17 Vo futileMainPV diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 6735 W: 1125 L: 1221 D: 4389
sprt @ 10+0.1 th 1 Don't futile prune main PV...
09-08-17 An noisyResearch5A diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 18560 W: 3315 L: 3361 D: 11884
sprt @ 10+0.1 th 1 Final take. Mix with a good local version of ttRootNode.
09-08-17 sg king_distance_history diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 9472 W: 1615 L: 1700 D: 6157
sprt @ 10+0.1 th 1 Introduce new history table based on moving piece and the distance to opponent king. Use them for quiet move sorting. (Park with prio -1).
09-08-17 pr ps_optimize diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 7304 W: 1283 L: 1377 D: 4644
sprt @ 10+0.1 th 1 test value convergence for 10+.1 games.
09-08-17 sg lmr_mc_depth diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 18696 W: 3285 L: 3331 D: 12080
sprt @ 10+0.1 th 1 Variation of green STC patch
09-08-17 Ro TestPTune diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 9007 W: 1523 L: 1649 D: 5835
sprt @ 10+0.1 th 1 Testing SPSA values "as is" (fixed benches)
09-08-17 sg move_order diff
LLR: -2.94 (-2.94,2.94) [0.00,5.00]
Total: 8422 W: 1439 L: 1528 D: 5455
sprt @ 10+0.1 th 1 Also use opponent king distance in LMR
09-08-17 be std_sort diff
LLR: -0.19 (-2.94,2.94) [-3.00,1.00]
Total: 1131 W: 197 L: 209 D: 725
sprt @ 10+0.1 th 1 Use stl for sorting, as baseline
05-08-17 II tmm_simple'' diff
LLR: -2.96 (-2.94,2.94) [0.00,4.00]
Total: 92716 W: 16496 L: 16341 D: 59879
sprt @ 40/10+0.1 th 1 A check whether we have a good constant for increment usage in x/y+z TC, take 1
09-08-17 sg lmr_mc_depth diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 21555 W: 2716 L: 2765 D: 16074
sprt @ 60+0.6 th 1 LTC: Combine with another of my yellow lmr tweak. Retry with now fixed formula (i inserted a term at the wrong place)
05-08-17 Ro PTune diff
29351/30000 iterations
60000/60000 games played
60000 @ 30+0.3 th 1 Pawn tuning for latest 2 patches
07-08-17 ia lever_count2 diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 85946 W: 15608 L: 15368 D: 54970
sprt @ 10+0.1 th 1 lever count. Take 2
09-08-17 pr ps_sort_custom diff
LLR: -2.95 (-2.94,2.94) [0.00,5.00]
Total: 6840 W: 1202 L: 1298 D: 4340
sprt @ 10+0.1 th 1 stage 1: optimizing sort limits for 10+0.1 games.
07-08-17 ia tune_pawn_file diff
24464/25000 iterations
50000/50000 games played
50000 @ 10+0.1 th 1 tune per file pawn eval arrays
07-08-17 ia sort_captures diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 78280 W: 14190 L: 13983 D: 50107
sprt @ 10+0.1 th 1 sort captures with same MVV/rank by LVA
08-08-17 Vo skipResearch diff
LLR: -2.96 (-2.94,2.94) [0.00,5.00]
Total: 56529 W: 10186 L: 10072 D: 36271
sprt @ 10+0.1 th 1 Don't skip early pruning if we need to do a research...